Welcome to Text-To-Video-Finetuning Discussions!

[ExponentialML]

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

• Ask questions you’re wondering about.
• Share ideas.
• Engage with other community members.
• Welcome others and are open-minded. Remember that this is a community we
  build together 💪.

To get started, comment below with an introduction of yourself and tell us about what you do with this community.

[Cubey42]

{'variance_type'} was not found in config. Values will be initialized to default values.
Traceback (most recent call last):
File "H:\AI video\Text-To-Video-Finetuning\train.py", line 488, in
main(**OmegaConf.load(args.config))
File "H:\AI video\Text-To-Video-Finetuning\train.py", line 226, in main
noise_scheduler, tokenizer, text_encoder, vae, unet = load_primary_models(pretrained_model_path)
File "H:\AI video\Text-To-Video-Finetuning\train.py", line 75, in load_primary_models
text_encoder = CLIPTextModel.from_pretrained(pretrained_model_path, subfolder="text_encoder")
File "C:\Users\Lucas\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\modeling_utils.py", line 2274, in from_pretrained
raise EnvironmentError(
OSError: Error no file named pytorch_model.bin, tf_model.h5, model.ckpt.index or flax_model.msgpack found in directory ./models/model_scope_diffusers/.

I only downloaded the fp16 models, do I need the full models for this to work?

[Cubey42]

okay it was just an extension of my previous issue, once I copied the whole directory, it worked. also wow even just from my test im getting some scary good results. insane what this is capable of

[ExponentialML]

@Cubey42 Glad to hear you got it working!

[sergiobr]

I got this on starting also. Now it's working fine.
Is there a way to git clone huggingface repo including all models?

[ExponentialML]

@sergiobr

This should do the trick. I'll add it to the readme.

git lfs install
git clone https://huggingface.co/damo-vilab/text-to-video-ms-1.7b

[sergiobr]

Great I knew there were a command for it. Thank you.